$ loading_
Review SIU indicators in claims and return explainable evidence with human review guidance.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "fraud-flag" skill from askskill: 1. Download https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/solutions/claims-processing/manual/skills/aibast_fraud-flag_03/SKILL.md 2. Save it as ~/.claude/skills/aibast_fraud-flag_03/SKILL.md 3. Reload skills and tell me it's ready
Using the provided synthetic claim records, identify which claim reaches the SIU review threshold. Separate observed evidence, calculated or heuristic output, and recommended next steps. Do not declare fraud.
Returns the claim ID, source-backed evidence, heuristic assessment notes, and recommended next steps requiring authorized human review.
Review record CLM-2025-7003 only. Summarize its SIU indicators from the current snapshot and clearly separate factual evidence from inferred conclusions. State that this is not legal, insurance, or financial advice.
Produces an indicator review for the specified record with evidence separation, disclaimers, and required human review.
Based on the given claim report scope, generate an SIU indicator review response: do not substitute a different record scope, do not perform any external action, and state that no approval, payment, or account change occurred.
Gets a structured response covering review scope, evidence sources, a no-external-action statement, and required human approval.
When claims staff need to answer whether a claim meets the SIU review threshold, this skill helps organize explainable risk indicators from the record. It focuses on evidence-backed review rather than declaring fraud.
SIU investigators can use it to separate observed facts, heuristic judgments, and proposed next steps for authorized human review. It also requires clearly stating that no external actions were taken.
The document describes an SIU indicator review skill used in a synthetic claims-processing pilot. Its purpose is to surface explainable fraud-related indicators without declaring fraud or changing coverage. The procedure requires using the exact fictional record scope, returning source-backed evidence from the synthetic snapshot, separating observed facts from heuristic output and next steps, and including disclaimers, a no-external-action statement, and the required authorized human review.
Surfaces explainable fraud indicators without declaring fraud or changing coverage.
Persona: SIU Investigator
Prompt: Which claim crosses the SIU review threshold, and does that prove fraud?
Expected synthetic evidence: CLM-2025-7003, SIU Referrals.
No. The documentation says it surfaces explainable fraud indicators without declaring fraud or changing coverage.
The output should separate observed evidence, calculated or heuristic output, and proposed next steps. It should also state that it is not legal, regulatory, insurance, or financial advice, and that no approval, payment, or external action occurred.
It requires using the exact synthetic record or report scope provided, without substituting another record, and returning source-backed evidence from the synthetic operating snapshot.
Identify at-risk client accounts and suggest recovery actions with risk context.
Extract and summarize medication inventories from packaged synthetic clinical evidence.
Summarize recovery conversion metrics with benchmarks, forecasts, and assumptions.
Drafts a consent-aware recovery outreach sequence without sending or scheduling.
Compare packaged satisfaction and NPS trends to identify declining accounts.
Extract a source-coded problem list from synthetic clinical evidence deterministically.
Checks claim-file readiness and compiles evidence for human review.
Triages fictional insurance claims and summarizes key intake details and priority.
Prepare reviewable retention and service-recovery options for churn-related customer cases.
Prioritize churn reviews with transparent evidence and suggested next steps.
Preview and coordinate vendor procurement for insurance claims investigations and reviews.
Estimate policy-term settlement, coverage limits, and deductibles for claims processing.